precocious s-phase entry in budding yeast prolongs replicative … · copyright 2002 by the...

Copyright 2002 by the Genetics Society of America

Precocious S-Phase Entry in Budding Yeast Prolongs Replicative State andIncreases Dependence Upon Rad53 for Viability

Julia M. Sidorova1 and Linda L. Breeden

Fred Hutchinson Cancer Research Center, Seattle, Washington 98109

Manuscript received August 30, 2001Accepted for publication November 5, 2001

ABSTRACTPrecocious entry into S phase due to overproduction of G1 regulators can cause genomic instability.

The mechanisms of this phenomenon are largely unknown. We explored the consequences of precocious Sphase in yeast by overproducing a deregulated form of Swi4 (Swi4-t). Swi4 is a late G1-specific transcriptionalactivator that, in complex with Swi6, binds to SCB elements and activates late G1-specific genes, including G1cyclins. We find that wild-type cells tolerate Swi4-t, whereas checkpoint-deficient rad53-11 cells lose viabilitywithin several divisions when Swi4-t is overproduced. Rad53 kinase activity is increased in cells overproduc-ing Swi4-t, indicating activation of the checkpoint. We monitored the transition from G1 to S in cells withSwi4-t and found that there is precocious S-phase entry and that the length of S phase is extended.Moreover, there were more replication intermediates, and firing of at least a subset of origins may havebeen more extensive in the cells expressing Swi4-t. Our working hypothesis is that Rad53 modulates originfiring based upon growth conditions to optimize the rate of S-phase progression without adversely affectingfidelity. This regulation becomes essential when S phase is influenced by Swi4-t.

PROPER integration of cell cycle transitions with investigate these mechanisms using budding yeast as amodel system.DNA metabolism is crucially important for cell sur-

Swi4 is a late G1-specific transcriptional activator, which,vival and error-free propagation of a cell’s genetic mate-in complex with Swi6, binds to SCB elements (Menden-rial. Cells that are unable to adjust the cell cycle clockhall and Hodge 1998) in the promoters of numerousupon receiving DNA damage are known to lose viabilitylate G1-specific genes (Iyer et al. 2001). The activity ofand/or compromise the fidelity of genetic transmissionthe Swi4/Swi6 complex gives rise to the concerted burst(Foiani et al. 2000). In the absence of any genotoxicof transcription of the G1 cyclin genes CLN1 and CLN2treatment, genomic stability can be negatively affectedand of a number of other genes, which promote theby relaxation of the control over the transition betweentransition to S phase (Nasmyth and Dirick 1991; OgasG1 and S phases. For example, forced entry of some qui-et al. 1991). Swi4-t, a C-terminally truncated and stabi-escent cells into S phase upon c-myc or cyclin E overpro-lized form of Swi4, is unable to interact with Swi6 andduction is known to result in genomic rearrangementscannot form the Swi4/Swi6 complex (Andrews and(Felsher and Bishop 1999; Mai et al. 1999; Spruck et al.Moore 1992; Primig et al. 1992; Sidorova and Breeden1999). Loss of Rb or ectopic expression of cyclin D1 in1993). However, unlike the full-length Swi4 (Baetz andtissue culture cells can increase gene amplification fre-Andrews 1999), the truncated Swi4-t is capable of bind-quencies (Donehower 1997). In budding yeast, inacti-ing to promoters independently of Swi6 (Andrews andvation of the B cyclin inhibitor SIC1, which causes earlierMoore 1992; Primig et al. 1992; Sidorova and Breedenactivation of the S-phase cyclin/CDK complexes and1993), and when overexpressed from the GAL promoterearlier entry into S phase (Schwob et al. 1994), leads(GAL::SWI4-t), it deregulates the transcription of targetto an elevated rate of chromosome loss (Nugroho andgenes (Breeden and Mikesell 1994). Swi4-t can causeMendenhall 1994). Overexpression of G1 cyclins CLN1a precocious entry into S phase similar to the checkpointor CLN2, whose levels are rate limiting for the G1-to-Smutant of Rad53 in cells that receive DNA damage intransition, is synthetically lethal with mutations in theG1 (Sidorova and Breeden 1997).

DNA damage checkpoint kinase gene MEC1 and causesRad53, a kinase conserved from yeast to humans, is in-

an increased level of chromosome loss (Vallen and volved in coordinating DNA metabolism with cell cycleCross 1995, 1999). The mechanisms of genomic insta- transitions (Weinert 1998; Foiani et al. 2000). In thebility induced by relaxation of control over the G1- presence of DNA damage in G1 and G2, Rad53 is mostto-S transition remain largely unknown. We sought to likely recruited to the damage recognition and pro-

cessing complex (Foiani et al. 2000) via Mec1-phosphor-ylated Rad9 (Emili 1998; Sun et al. 1998; Vialard et al.1998; Durocher et al. 1999). Rad53 is phosphorylated1Corresponding author: Fred Hutchinson Cancer Center, 1100 Fair-

view Ave. N, Seattle, WA 98109. E-mail: [email protected] and activated in a Mec1 kinase-dependent manner (Navas

Genetics 160: 123–136 ( January 2002)

124 J. M. Sidorova and L. L. Breeden

trp1 ade5. Mating tester strains BY26 and BY27 are MATa his1et al. 1996; Sanchez et al. 1996; Sun et al. 1996). Activatedand MAT� his1, respectively.Rad53 and Mec1 phosphorylate specific targets, which

The plasmid pBD1168 is a YCp50 vector with GAL::SWI4-tcan delay cell cycle transitions and induce DNA repair and was described previously (Sidorova and Breeden 1993).(Allen et al. 1994; Sidorova and Breeden 1997; Hu- pBD1411 and pBD1948 were described before (Tyers et al.

1993; Epstein and Cross 1994). The plasmid pBD2385 hasang et al. 1998; Pellicioli et al. 1999; Sanchez et al.been described previously (Sidorova and Breeden 1993) and1999; Bashkirov et al. 2000; Kihara et al. 2000). Rad53is a Ycp50 with the GAL promoter-driven SWI4 gene. Theis also important for monitoring DNA damage and nu-plasmid pBD2972 is a kind gift of Dr. S. Elledge and contains

cleotide shortages within S phase (Paulovich and Hart- the RNR1 open reading frame under the control of the GALwell 1995; Navas et al. 1996). Rad53 prevents firing of promoter on a 2-�m TPR1 shuttle vector (Desany et al. 1998).

Growth conditions: All rich (YEP) and minimal (YC) mediaorigins when DNA replication is encumbered by dam-and growth conditions were as described before (Breedenage (Santocanale and Diffley 1998; Shirahige et al.and Mikesell 1991). Cultures used for elutriation were grown1998; Tersero and Diffley 2001) and potentially stabi-in YC-ura glucose media and then inoculated into YEP media

lizes stalled replication forks (Lopes et al. 2001). Rad53 with 2% raffinose and grown to OD660 � 1.5–1.8. To inducealso blocks premature metaphase-to-anaphase transi- Swi4-t, galactose was added 80–100 min before the zero time

point of the experiment (this induction time included thetion in cells with incompletely replicated DNA (Allentime in the elutriator). The zero time point was the time whenet al. 1994). Finally, Rad53 is an essential gene (Zhengthe unbudded eluted cells were placed in the 30� incubator.et al. 1993) whose role in the unperturbed cell cycleThis time was ample to fully induce Swi4-t overexpression,

is most likely involved in S-phase progression control which takes �30 min ( J. Sidorova, unpublished results), and(Desany et al. 1998; Dohrmann et al. 1999). Along with ensured that these cells had high levels of Swi4-t from the very

beginning of G1 phase. For �-factor synchrony experiments, aMec1, Rad53 may act during S phase to remove theculture at OD660 � 0.2 was typically arrested by incubationSml1-mediated inhibition of ribonucleotide reductasewith 5 mg/liter of �-factor for 1 hr 45 min. Cells were released(Zhao et al. 1998).from the arrest by filtration or by addition of Pronase E (Sigma,

In this study, we demonstrate that Swi4-t overexpres- St. Louis) to a final concentration of 10 mg/liter. dbf4-1 strainssion can cause precocious entry into S phase in the absence were arrested at 37� for 2.5–3 hr and released into the cellof exogenous DNA damage. We also show that Swi4-t cycle by shifting back to 25�.

FACS analysis: FACS analysis was done exactly as describedoverexpression results in a loss of viability in combina-before (Sidorova and Breeden 1997) except that cells were ana-tion with an allele of RAD53 that is defective in check-lyzed on a Calibur Analyser (Becton Dickinson, San Jose, CA).point function (rad53-11). Analysis of S phase in RAD Elutriation: Elutriation was performed in a J-6B centrifuge

and rad53-11 strains overexpressing Swi4-t suggests that (Beckman, Fullerton, CA) using a JE-5.0 rotor and a 40-ml cham-Swi4-t prolongs the replicative state of cells and may ber (Beckman). The chamber was loaded with cells at 3500 rpm

and 28 ml/min flow rate. Flow rate was then increased to 35increase the frequency of replication initiation. We pro-ml/min and the chamber was equilibrated with fresh media.pose that excessive origin firing can result in stalling ofSmall G1 cell fractions were harvested in fresh media by fur-forks due to depletion of resources such as dNTPs or ther increasing the flow rate from 35 ml/min to 55 to 60 ml/

histones. In the absence of Rad53, stalled forks are not min in 3- to 5-ml/min increments. The size of cells in thesestable enough to resume replication when resources are fractions was determined using a Z2 Particle Count and Size Ana-

lyzer and the data were analyzed using the AccuComp versionreplenished, and there is no signal to inhibit further origin2.01 software (Beckman Coulter, Miami).firing. Combined, these deficiencies lead to lethality. In

Nocodazole execution point measurements: To determineother words, during a normal S phase, Rad53 servesthe nocodazole execution point, elutriated G1 cells were al-

as a pacemaker, coordinating S-phase progression with lowed to progress through the cell cycle. Aliquots (5 ml) of thisgrowth conditions. By preventing excessive origin firing, culture were taken every 10 min, transferred to culture tubesit minimizes the effects of precocious S-phase entry im- with nocodazole (final concentration 12 mg/liter), and incu-

bated at 30� for 2 hr on a roller drum. After 2 hr these cells wereposed by hyperactive Swi4-t.treated with 1% sodium azide and sonicated and the cell num-bers were counted. As expected, cells treated with nocodazolebefore the first anaphase were arrested in the first cycle (JacobsMATERIALS AND METHODSet al. 1988). Cells that had passed this transition divided onceand arrested in metaphase of the second cycle. Thus, theStrains and plasmids: The yeast strain BY2006 MATa ura3 leu2final cell concentration after the nocodazole treatment wastrp1 his3 has been described (Sidorova and Breeden 1997).dependent upon the proportion of cells that had traversedBY2390 is a ura3 derivative of BY2007 MATa ura3 leu2 trp1 his3the first anaphase and become nocodazole resistant at the timerad53-11::URA3 (Sidorova and Breeden 1997). BY2912 andpoint when the aliquot was taken. Percentage of nocodazole-BY2913 were derived from BY2006 and BY2390, respectively,resistant cells at each time point was calculated using theseby integration of Escherichia coli dam methylase on a pRS305 vec-cell concentration data. In addition, aliquots of cells weretor into the endogenous LEU2 locus (Friedman et al. 1995).taken every 10 min from the untreated growing culture. TheseBY2887 and BY2888 are isogenic to BY2006 and BY2390, re-samples were also arrested by 1% sodium azide, and cell con-spectively, except they carry a rad52::LEU2 disruption. BY2914centration and proportion of budded cells were determinedis MATa ura3 lys2 leu2 trp1 pep4::HIS3 prb1�1.6R with the en-in each sample.dogenous RAD53 gene tagged with the HA tag (Emili 1998).

MAT locus heterozygosity loss measurements: These mea-BY2226 MATa ura3 leu2 trp1 his3 mec1-1::HIS3 and BY2227surements were done essentially as described (Rose et al. 1990),MATa ura3 leu2 trp1 his3 rad9�::LEU2 were described before

(Sidorova and Breeden 1997). BY479 is MATa dbf4-1 ura3 taking advantage of the fact that diploid cells that lose one

125Dependence on RAD53 in Yeast With Early S Onset

of the MAT alleles will mate as haploids. RAD/rad53-11 hetero-zygous diploids were constructed by mating BY2006 MATaura3 leu2 trp1 his3 (Sidorova and Breeden 1997) and BY2918MAT� ura3 leu2 his3 ade rad53-11::ura3::LEU2. Diploids weretransformed with the empty vector pYES2 or with the GAL::SWI4-t pBD1168 plasmid and were grown in YC-ura media with2% raffinose and 2% galactose for about eight generations. Ofthese cells, 5 � 106 were mated with the equal number of cellsof either one of the mating tester strains, BY26 or BY27. Uponplating, mated cells gave rise to prototrophic colonies. A sepa-rate control was included with only RAD/rad53-11 and nomating tester strains. Neither RAD/rad53-11 diploids nor mat-ing tester strains alone gave rise to prototrophic colonies ata measurable frequency (�10�6). Loss of heterozygosity wascalculated as the number of mating events per milliliter ofculture divided by the number of viable cells per milliliter ofculture, and it was typically �10�4. The average numbers andaverage deviation of five measurements are reported.

Rad53 kinase assay: Immunoprecipitation of HA-Rad53 andkinase assays were done as described before (Sidorova andBreeden 1997) with 12CA5 mouse monoclonal antibodies tothe HA epitope, except that no exogenous substrate was addedto the kinase reaction.

Hemi-methylation analysis: Hemi-methylation analysis wasdone as described (Friedman et al. 1995, 1997). A constitutivelyexpressed E. coli dam methylase gene on a pRS305 plasmid (Fried-man et al. 1995) was integrated into the endogenous LEU2 locusof yeast strains. These strains did not show a growth disadvantagecompared to the parental strains, and GAL::SWI4-t overexpres-sion had the same phenotype in the methylated strains as inthe parental strains. DNA was digested overnight with DpnIand EcoRI and loaded onto 0.8% agarose gels. Southern blot-ting and hybridization with the ARS607 probe were performedon GeneScreen membranes (New England Nuclear, Boston)according to the manufacturer’s recommendations. The probe

Figure 1.—Overexpression of Swi4-t is lethal in rad53-11 orfor the ARS607 region maps to chromosome VI coordinatesmec1-1 mutants, but not in rad9� mutants. (A) Wild-type (BY2006),from 202,454 to 203,480 (http://www-genome.stanford.edu/rad53-11 (BY2390), mec1-1 (BY2226), and rad9� (BY2224) strainsSaccharomyces/) downstream of the ARS607 ARS consensus se-were transformed with the empty vector (Ycp50), GAL::SWI4-tquence. In a hemi-methylated or unmethylated DNA, this probeplasmid (pBD1168), GAL::CLN1 plasmid (pBD1141), or GAL::preferentially hybridizes to a DpnI-resistant EcoRI-digested frag-CLN2 plasmid (pBD1948) and streaked onto selective platesment of �3.0 kb (DpnIR). Full methylation of DNA allows thiswith galactose. (B) The rad53-11 strain carrying GAL::SWI4-tfragment to be further digested by DpnI. In this case, thewas transformed with either the empty vector p414GAL1 orprobe hybridizes mostly to a DpnI-sensitive (DpnIS), EcoRI/the GAL::RNR1 plasmid (pBD2972). The strains were grownDpnI-digested fragment of �1.0 kb.on selective media plates with galactose. (C) 105, 0.33 � 105,Two-dimensional gel electrophoresis: We followed the pre-and 104 of cells of the wild-type (BY2006), rad52� (BY2887),viously described procedures for the DNA isolation and two-rad53-11 (BY2390), and rad53-11 rad52� (BY2888) strains car-dimensional gel electrophoresis of replicative intermediatesrying GAL::SWI4-t plasmid (pBD1168) were spotted onto selec-(Friedman and Brewer 1995). The isolated DNAs were di-tive media plates with galactose.gested with NheI for rARS replicative intermediates and re-

solved as follows. The first dimension separation was through0.4% agarose at 1 V/cm for 20 hr at room temperature, and eliminate the methyl methanesulfonate-dependent de-the second dimension was run in 1% agarose at 5 V/cm for

lay of the G1/S transition (Sidorova and Breeden 1997).5 hr at 4�. To visualize ARS608, DNAs were digested with XhoITo explore whether combining these two genetic deter-and resolved in 0.4% agarose at 1 V/cm for 24 hr in the first

dimension and in 0.8% agarose at 2.7 V/cm for 15 hr at 4�. minants in one background will result in a completeThe probe for rARS maps to chromosome XII coordinates elimination of this delay, we introduced three different460,050–460,777. The probe for ARS608 maps to chromosome SWI4 alleles under control of the GAL1 promoter intoVI coordinates 213,306–214,317.

a rad53-11 strain. These alleles encoded wild-type Swi4and two of its truncated derivatives lacking 140 or 280amino acids from the C terminus. We found that overex-RESULTSpression of the 280-amino-acid truncation Swi4-t from

Cells overproducing Swi4-t require a functional S-phase the GAL promoter (GAL::SWI4-t) severely abrogated thecheckpoint for viability: We have shown before that the rad53-11 strain’s viability even in the absence of exoge-checkpoint-deficient allele of RAD53 (rad53-11) or over- nous DNA damage (Figure 1A). These cells typicallyproduction of a C-terminally truncated form of Swi4 formed micro-colonies of up to 20 to 40 cells and divi-

sion was arrested within 48 hr of growth on galactose.(Swi4-t) both significantly reduce but do not completely


Neither the full-length Swi4 nor a shorter truncation of140 amino acids compromised the strain’s viability (datanot shown).

Vallen and Cross (1995, 1999) have previously re-ported that high levels of either of the G1 cyclins, Cln1or Cln2, are lethal in combination with the checkpoint-deficient mec1 allele in the BF264-15D background. SinceCLN1 and CLN2 levels increase upon induction of Swi4-t(Breeden and Mikesell 1994), we examined whetherthese cyclins were sufficient to cause lethality in theA364a rad53-11 strain and whether GAL::SWI4-t was le-thal to a mec1 strain in the A364a background. Overex-pression of Swi4-t did lead to poor viability with mec1-1.However, GAL::CLN1 or GAL::CLN2 did not abolish col-ony formation in rad53-11 or mec1-1 strains (Figure 1A). Figure 2.—GAL::SWI4-t induces Rad53 kinase activity. AThese results suggest that the lethality caused by Swi4-t wild-type strain (BY2914) with a HA-tagged Rad53 was trans-in the A364a background cannot be attributed entirely formed with the empty vector pYES2 (top) or with the GAL::

SWI4-t plasmid (pBD1168, bottom). The cells were grown into G1 cyclin overproduction.selective media with raffinose and then diluted into fresh me-Mec1 and Rad53 are required for all three DNA dam-dia with galactose to induce Swi4-t expression. Aliquots wereage checkpoints in G1, S, and G2 phases of the cell cycle taken immediately before galactose was added and every hour

(Foiani et al. 2000). Rad9, on the contrary, is predomi- for 6 hr. Rad53 was immunoprecipitated with HA antibodiesnantly required for the G1 and G2 checkpoints (Navas and allowed to autophosphorylate in the presence of [-32PO4]

ATP. Products of the kinase reactions were resolved on anet al. 1996). We found that Swi4-t overexpression hadSDS polyacrylamide gel and autoradiographed. Positions ofno effect on the viability of a rad9� strain (Figure 1A).the modified and unmodified forms of Rad53 are marked byThe fact that Swi4-t is lethal in the mec1-1 and rad53-11 lines on the right.

but not in the rad9 background suggests that functionsof Mec1 and Rad53 that do not overlap with Rad9’sfunction are critical for protection against Swi4-t. Re- Swi4-t induces Rad53 kinase activity and speeds up

the G1/S transition: DNA damage or stalled replicationsponse to disruptions in DNA replication is one suchfunction (Foiani et al. 2000). forks cause modification and activation of the Rad53

kinase. This modification results in the appearance ofIf the lethal effect of Swi4-t on rad53 cells is associatedwith the way the S phase progresses in these cells, we would a low-mobility form of Rad53 on protein gels (Allen et

al. 1994; Sanchez et al. 1996; Sun et al. 1996). Figure 2anticipate that gene products important for properS-phase progression might have synthetic phenotypes shows the state of Rad53 immunoprecipitated out of wild-

type cells harboring vector control plasmid or GAL::SWI4-twhen overproduced in a GAL::SWI4-t rad53-11 strain.RNR1, the large subunit of ribonucleotide reductase and grown in galactose. An unmodified, high-mobility

form of Rad53 remains the predominant form of thisthat is an important and perhaps limiting componentof an S-phase cell, rescues the lethality of GAL::SWI4-t kinase in vector control cells throughout the time

course. In contrast, after 2 hr of galactose induction ofrad53-11 cells when overexpressed (Figure 1B). Interest-ingly, RNR1 is a target of Swi4 in vivo (Iyer et al. 2001) Swi4-t, there is a detectable accumulation of the low-

mobility form of Rad53. Since typically only the unmodi-and RNR1 is transcribed at a higher level when Swi4 orSwi4-t is overexpressed (data not shown). fied form of Rad53 is readily detectable during a normal

cell cycle (Sun et al. 1996; J. Sidorova, data not shown),The presence of Swi4-t could generate some kindof damage during replication, which could evoke the this increased presence of the low-mobility, modified

form of Rad53 upon induction of Swi4-t suggests thatRad53-dependent checkpoint and necessitate repair. In-deed, RAD52, which is critical for DNA recombination these cells are undergoing a Rad53-mediated check-

point response.and repair (Sung et al. 2000), has a synthetic phenotypewith rad53-11 GAL::SWI4-t. The rad52 deletion exacer- Swi4 is a critical activator of the G1/S transition, and

Swi4-t is a hyperactive form that deregulates transcrip-bates Swi4-t-induced lethality in rad53-11 cells while hav-ing no measurable effect on RAD53 cells (Figure 1C). tion of G1 cyclins (Breeden and Mikesell 1994). As

such, Swi4-t overproduction is expected to affect theIn addition, long-term Swi4-t overexpression results inan increase in chromosome instability as judged by the G-to-S transition. However, it is less evident that Swi4-t

may affect S-phase progression. We used centrifugal elu-loss of heterozygosity of the MAT locus. A RAD/rad53-11heterozygous diploid strain overexpressing Swi4-t shows triation to examine cell cycle progression in Swi4-t cells

with or without functional Rad53. Swi4-t was inducedabout a threefold higher frequency of allele loss eventsat the MAT locus than cells carrying the vector alone 120 min before the first measurement was taken. This

protocol enabled us to look at the immediate effects of(1.25 0.33 � 10�4 vs. 4.18 0.66 � 10�4).


Figure 3.—GAL::SWI4-t speeds up budding and S-phase onset in elutriated G1 cells. (A) Wild-type (BY2912) or rad53-11(BY2913) cells were transformed with GAL::SWI4-t (pBD1168) or empty vector (Ycp50) and grown in rich medium with raffinose,galactose was added for 20–40 min, and small G1 cells were harvested by centrifugal elutriation in fresh YEP media with raffinoseand galactose. Fractions of identical cell size distribution were collected and allowed to progress through the cell cycle. Sampleswere taken from these cultures every 10–15 min and the percentage of budding cells was determined. (B) Wild-type strain fractionsand rad53-11 strain fractions obtained by elutriation were subjected to FACS analysis. Mean cell volumes at the corresponding timepoints are indicated to the right of the profiles.

GAL::SWI4-t on the cell cycle. The fractions collected initiation of budding of the control vs. Swi4-t cells (Fig-ure 3A). As with control cells, the budding induced byfrom the elutriator were allowed to progress through

the cell cycle and sampled at 10-min intervals. Overex- Swi4-t was accompanied by an earlier entry into S phaseas determined by FACS (Figure 3B), indicating thatpression of Swi4-t did not change the kinetics of cell

growth in G1 (data not shown), but it caused cells to bud these events remain coupled in the Swi4-t cells. rad53-11cells carrying Swi4-t reproducibly exited G1 at a smallerat a smaller size than controls. For example, GAL:SWI4-t

cells achieved 50% budding at the average mean volume size than control cells. However, the volume differencebetween GAL::SWI4-t and vector cells was less dramaticof 17.4 fl, while the vector control cells were 50% bud-

ded at the average mean volume of 24 fl. Direct compari- than in the case of the wild-type cells (20 fl vs. 23 fl;see also Figure 3, A and B). We also found that elutriatedson of fractions of cells with identical starting size distri-

butions showed that there was a 25-min difference in the rad53-11 cells exhibited a less synchronous exit from G1


TABLE 1

Timing of the G1-to-S and metaphase-to-anaphase transitions

Starting 50% budding 50% nocodazole resistance Time betweenStrain mean volume (fl) (G1/S transition, min) (metaphase-to-anaphase transition, min) the transitions

RAD/vector 15.62 57 134 77RAD/GAL::SWI4-t 15.67 34 120 86rad53/vector 16.84 62 143 81rad53/GAL::SWI4-t 16.72 46 �200a �154

Cells of the genotypes indicated were grown in raffinose and then galactose was added 20 min prior to the elutriation. G1cells of the indicated size were harvested and their passage through the cell cycle was followed. Cell number and percentage ofbudded cells were determined at each time point. After 40 min, aliquots of each culture were removed every 10 min and in-cubated with 12 mg/liter of nocodazole for 2 hr. These cells were counted and the percentage of cells that had passed thenocodazole execution point and were able to divide once was calculated.

a rad53-11 GAL::SWI4-t cells did not exceed 50% nocodazole resistance within the time frame of the experiment.

than their wild-type counterparts. Budding was slower region can be followed as a transient state of hemi-methylation, because hemi-methylated DNA is resistantand 1N DNA persisted for a longer time in these cells

(Figure 3B). to cleavage by the DpnI restriction enzyme. To compareGAL::SWI4-t and vector cells, elutriated G1 cells of theComparison of budding indices of the synchronized

GAL::SWI4-t and vector cells indicates that GAL::SWI4-t two strains were followed as they progress through G1and S, but were arrested in anaphase by adding micro-cells may spend more time as budded cells (Figure 3A).

Moreover, the FACS data (Figure 3B) suggest that while tubule inhibitor nocodazole (Figure 4A). Alternatively,we employed �-factor synchronization (Figure 4, B–D).GAL::SWI4-t cells traverse the G1-to-S transition earlier

than controls, they spend a longer time completing S This method induces a higher degree of synchrony butthe first G1-to-S transition is very rapid.phase. In GAL::SWI4-t cells, the fraction of cells with 2N

DNA content gradually rises between 60 and 140 min Consistent with previous observations, in elutriated wild-type G1 cells arrested before anaphase by nocodazole,with no indication of progression into the next cell

cycle. In contrast, the control cells show no pausing at the onset of S phase and replication through the ARS607region as assayed by DpnI resistance occurred earlier inthe 2N or near 2N DNA stage and proceed rapidly

into the next cycle. To address this in more detail, we GAL::SWI4-t cells than in controls and just as efficiently,as indicated by the sharp increase of hemi-methylationobtained elutriated populations of cells with identical

starting size and monitored their progression between level (Figure 4A). However, despite the earlier onset ofreplication, the hemi-methylated state of the ARS607 locusG1 and G2 phases. Swi4-t cells budded 23 min earlier

than controls (Table 1). However, the execution of ana- persisted at a high level for a longer time in GAL::SWI4-t cells than in controls.phase, as measured by the acquisition of 50% resistance

to the microtubule-destabilizing drug nocodazole, oc- Using the �-factor synchronization we observed a rapidG1-to-S transition in both GAL::SWI4-t and vector cells (Fig-curred only 14 min earlier in Swi4-t cells compared to

controls (Table 1). Hence, it appears that Swi4-t cells ure 4B). This was expected because �-factor-arrestedcells exceed the critical size needed for the transitionspend more time between the end of G1 phase and

anaphase. rad53-11 cells overexpressing Swi4-t also expe- into S phase and thus exit G1 very quickly upon release.Nonetheless, the hemi-methylation assay showed thatrienced an extension of this interval (Table 1). rad53-11

GAL::SWI4-t cells were less efficient in completing the while both GAL::SWI4-t and vector control cells startedto accumulate newly replicated, hemi-methylated DNAcell cycle, since the number of nocodazole-resistant cells

never exceeded 50%. This suggests that these GAL::SWI4-t in the early ARS607 locus at about the same time,GAL::SWI4-t cells carried hemi-methylated DNA for acells do not undergo premature mitosis due to mutation

in RAD53. longer time than vector controls (Figure 4, C and D).In rad53-11 cells, induction of hemi-methylation was lowExtended replicative state in cells expressing Swi4-t:

The extension of the interval between G1 and anaphase due to a low basal level of methylation, precluding adefinitive interpretation of results.in cells overexpressing Swi4-t suggests that Swi4-t may

slow or impair DNA replication. To follow the replica- The hemi-methylation studies suggest that Swi4-t over-expression can result in both an earlier appearance andtion more directly, we employed hemi-methylation anal-

ysis (Friedman et al. 1997). We used the strains that a prolonged presence of nascent DNA in the ARS607region in wild-type cells. However, it is also possible thatconstitutively express E. coli dam methylase, which meth-

ylates both strands of yeast DNA in vivo. In strains car- the observed differences in hemi-methylation profilesare caused by an altered accessibility to dam methylaserying dam methylase, replication through a particular


Figure 4.—GAL::SWI4-t extends the interval during which nascent DNA can be detected by hemi-methylation analysis. (A)The wild-type (BY2912) strain transformed with the empty vector (Ycp50) or with GAL::SWI4-t (pBD1168) was elutriated asdescribed for Figure 3. Mean volumes of the starting cultures were 16.4 fl for both wild-type strains. These cultures were allowedto progress into S phase, and after 40 min nocodazole was added to the final concentration of 12 mg/liter to prevent escapeinto the second cell cycle. Samples were taken every 10 min to isolate DNA and methylation state of nascent DNA was followedby digesting with DpnI and EcoRI. The Southern blots of EcoRI, DpnI-digested DNA were hybridized with probes to the ARS607region and quantified. The DpnI resistance was determined as a ratio of the DpnI-resistant (DpnIR) to the sum of the DpnI-resistant and DpnI-sensitive (DpnIS) DNA. The DpnI resistance at time point 0 was set equal to 1 unit, so that the fold inductionof DpnI resistance could be compared for the indicated strains. (B–D) The wild-type (BY2912) strain transformed with the emptyvector (pBD1129) or with GAL::SWI4-t (pBD1168) was synchronized in late G1 by incubating with �-factor for 105 min. Galactosewas added 30 min before the release. Cells were released from the arrest by filtration and resuspended in fresh media withraffinose and galactose. Aliquots of cultures were taken at 10-min intervals and analyzed by FACS (B) and hemi-methylationanalysis (C). Southern blot hybridizations to ARS607 region DNA are shown (C), and quantitation of these data is shown in(D). Hemi-methylated DNA is resistant to DpnI digest, giving rise to a large EcoRI-EcoRI fragment (DpnIR), and fully methylatedDNA is sensitive to DpnI digest, giving rise to a smaller EcoRI-DpnI-digested (DpnIS) fragment.

after one round of replication, for example, due to a of the hemi-methylated state in DNA upon Swi4-t overex-pression correlates with the prolonged replicative statechange in the nascent chromatin structure.

Swi4-t overexpressing cells display more replicative and to assess replication in rad53-11 cells, we employedtwo-dimensional (2D) gel electrophoresis (Figure 5A;intermediates during S phase: To test whether extension


Figure 5.—GAL::SWI4-t extends the window of activity of the ribosomal ARS in wild-type cells. (A) Diagram depicting themigration patterns of replication intermediates of ribosomal DNA through 2D gels. (B) A wild-type (BY2006) strain transformedwith the empty vector (Ycp50) or with GAL::SWI4-t (pBD1168) was synchronized with �-factor. Galactose was added 30 min be-fore the release. Cells were released from the arrest by adding Pronase E (Sigma) to the final concentration of 10 mg/liter andaliquots of cultures were taken at specified times for FACS analysis (B) and 2D gel analysis (C). The DNAs were digested with theNheI restriction enzyme and resolved in two dimensions. Gels were subjected to Southern blotting and hybridization and probedfor the ribosomal ARS locus as specified in materials and methods.

Friedman and Brewer 1995). In these gels, bubble- (Figure 5C). In addition to bubble- and Y-shaped mole-cules, two other types of replicative intermediates canshaped intermediates arise from origin firing within the

given DNA fragment. Y-shaped intermediates are gener- be detected in rDNA during replication. Leftward-mov-ing forks that stall at the replication fork barrier (RFB)ated either when forks are initiated distally and passively

replicate through the monitored fragment or when the show up as a dot of increased intensity on the Y arc (Lins-kens and Huberman 1988; Brewer et al. 1992). In addi-replication bubble is positioned asymmetrically within

the fragment and one fork completes its replication tion, at later stages of replication, converging forks formX-shaped, Holliday junction-like intermediates, whichbefore the other. To quantify replicative intermediates,

signal intensities from bubbles or Ys can be normalized migrate separately from simple Ys (Figure 5, A and C).Thus, the rDNA cluster provides an opportunity to mon-to that of the double-stranded DNA, which migrates as

a spot in front of the replicating molecules (1N DNA, itor accumulation of the regular early (bubble) and late(Y) intermediates, including the stalled and convergedFigure 5A; Ivessa et al. 2000).

Using this assay, we first followed replication through replicative forks.To follow replication, cells were synchronized in latethe ribosomal gene cluster on chromosome XII, which

spans 1000–2000 kb of 9-kb repeats of rDNA each contain- G1 by �-factor. The highest levels of replicative interme-diates in the wild-type cells were detectable at 50 mining an ARS sequence (rARS). Replication intermediates

of the rDNA cluster are detectable throughout S phase (Figure 5, B and C). At this time and later, we could


Figure 6.—GAL::SWI4-t increases the levels of replicative intermediates detectable in rad53-11 cells. A rad53-11 (BY2390) straintransformed with the empty vector (Ycp50) or with GAL::SWI4-t (pBD1168) was treated as described for the wild-type strain inthe legend to Figure 5. Samples of cells were taken every 10 min to determine DNA content by FACS. (A) Note that GAL::SWI4-tcells are slightly ahead of the vector controls in S-phase progression. (B) DNAs isolated from these cultures at the times specifiedwere digested with the NheI restriction enzyme and resolved in two dimensions as in Figure 5. The arrow points to the X-shapedintermediates prominent in the rad53-11 GAL::SWI4-t cells. (C) A shorter exposure of the two-dimensional gel from an independentexperiment is shown to compare the abundance of the RFB-stalled forks in the rad53-11 vector vs. rad53-11 GAL::SWI4-t cells.Positions of RFB-stalled forks and of 1N DNA are marked by arrows. (D and E) Autoradiograms of rARS replicative intermediatesof the rad53-11 BY2390 strain with the empty vector or GAL::SWI4-t were quantified, and the signals in bubbles and Ys (D) or inthe RFB-stalled fork (E) intermediates were normalized to the signal in 1N DNA and plotted. The average of two measurementsis presented.

detect higher levels of Y intermediates in GAL::SWI4-t titation of these gels confirms the observation that thepresence of GAL::SWI4-t results in about twice as manycells compared to controls, suggesting that replicative

intermediates persist for a longer time in these cells. In Ys and RFB-stalled forks (Figure 6, D and E). Thus, thetwo-dimensional gel analysis suggests that replicative forksrad53-11 cells, the effect of Swi4-t overexpression was

more pronounced (Figure 6). The rad53-11 GAL::SWI4-t are present on rDNA for an extended time in GAL::SWI4-tcells. In addition, the levels of bubble intermediatescells maintained higher levels of bubbles, Ys, and stalled

and converged forks than did the rad53-11 vector con- evident in a fixed amount of DNA are two- to threefoldhigher in rad53-11 GAL::SWI4-t cells than in controls attrols (Figure 6B). Figure 6C shows a repeat of this experi-

ment with a lighter exposure so that the levels of the the earliest time point (see Figure 6, B and D). Thesedata suggest that the ribosomal ARS fires more fre-RFB-stalled forks in rad53-11 GAL::SWI4-t vs. rad53-11

vector cells can be compared (marked by arrows). Quan- quently when GAL::SWI4-t is present.


rad53-11 and GAL::SWI4-t may influence origin firing S phase (Friedman et al. 1997; Yamashita et al. 1997).As seen in Figure 7A, both in RAD and in rad53-11 cells,frequency: We next followed replication of a single copy

origin, ARS608, whose firing frequency is variable and GAL::SWI4-t extended the window of time during whichlimited to a narrow window of time in the first half of the large bubble intermediates (marked by arrows) were

detectable in ARS608 DNA by at least 10 min. Y interme-diates in GAL::SWI4-t cells were also more prevalent latein S phase.

Interestingly, we also observed a difference betweenthe relative amounts of Y and bubble intermediates inRAD and rad53-11 cells. An important difference betweenthe rARS and ARS608 is the fact that the latter is a single-copy ARS and it is replicated either actively, if it fires,or passively from the nearby ARS607. Thus, the bubble-to-Y ratio is a good indicator of the firing efficiency ofARS608 (Friedman et al. 1997; Yamashita et al. 1997).As Figure 7B shows, the maximal detectable firing fre-quency of ARS608 was reproducibly increased abouttwofold in rad53-11 cells. The quantification also showedthat in both RAD and rad53-11 cells, Swi4-t overexpres-sion did not lead to a significant increase in the ARS608bubble-to-Y ratio during the peak activity of this ARS.Overall, the 2D analysis indicates that both rad53-11 andGAL::SWI4-t contribute to the increase in origin firingalthough these contributions are manifested differently.While rad53-11 shifts the distribution toward early inter-mediates, GAL::SWI4-t extends the window of time dur-ing which the intermediates are detectable.

To further address the effects of Swi4-t, we asked whe-ther it can suppress replication initiation mutations. DBF4is an essential gene required for initiation but not elonga-tion of replication throughout S phase (Johnston andThomas 1982; Bousset and Diffley 1998; Donaldsonet al. 1998). dbf4-1 is a temperature-sensitive mutation,which reduces the Dbf4/Cdc7 complex kinase activity(Kihara et al. 2000) and prevents replication initiation atthe nonpermissive temperature (Johnston and Thomas1982). We found that Swi4-t, but not the full-length Swi4,improved the growth of the dbf4-1 strain at the semi-permissive temperature of 33� (Figure 8A). We thenarrested the GAL::SWI4-t and vector cells at the begin-ning of S phase by shifting them to 37�, induced Swi4-texpression, and then shifted them back to the permis-sive temperature. FACS analysis showed that the dbf4-1

Figure 7.—GAL::SWI4-t extends the window of time duringwhich ARS608 replicates. (A) The wild-type (BY2006) or rad53-11(BY2390) strains with the empty vector or GAL::SWI4-t pre-sented in Figures 5 and 6 were used to examine ARS608replication. DNAs from specified time points were digestedwith the XhoI restriction enzyme and resolved in two dimen-sions as specified in materials and methods and probedwith ARS608 locus DNA. (B) Three to four blots of ARS608replicative intermediates, taken from the times when ARS608showed the most activity, were quantified using PhosphoImagerfor each strain and the ratio of the signals in the full arc ofbubble intermediates and the full arc of Y intermediates wasaveraged and plotted.


and ARS608 DNA in cells carrying GAL::SWI4-t suggeststhat some replicative forks in these regions were slowingor stalling. Another interpretation of this result is thatSwi4-t may cause an extension of the window of timewithin which origins fire and/or an increase in the fre-quency with which they fire. Support for the notion thatSwi4-t may increase efficiency of at least some origins isrendered by the fact that Swi4-t overexpression partiallysuppresses the temperature sensitivity of the dbf4-1 strainand speeds up the course of S phase in dbf4-1 cells. Dbf4is involved in initiation of DNA replication (Johnstonand Thomas 1982), and one may expect that the con-sequences of partial Dbf4 inactivation could be counter-acted by upregulating origin firing.

The cellular response to Swi4-t involves Rad53. Rad53is required for viability, and when hyperactivated in re-sponse to DNA damage, it slows down S-phase progres-sion by inhibiting origin firing (Santocanale and Diff-

Figure 8.—GAL::SWI4-t partially suppresses temperature ley 1998; Shirahige et al. 1998; Tersero and Diffleysensitivity of the dbf4-1 strain. (A) A dbf4-1 (BY479) strain trans- 2001). Swi4-t overexpression leads to the activation offormed with the empty vector (Ycp50), GAL::SWI4 (pBD2385),

the Rad53 kinase. rad53-11 checkpoint mutant cells areor GAL::SWI4-t (pBD1168) was streaked onto selective medianot able to tolerate overexpression of Swi4-t and loseplates with galactose and grown at 33�. (B) The dbf4-1 (BY479)

strain with the empty vector (Ycp50) or with GAL::SWI4-t (pBD- viability within three to five divisions. These observations1168) was grown at 25� in YEP media with raffinose and ar- suggest that Swi4-t overexpression invokes a checkpointrested before S phase by incubation at 37� for 2.5 hr. Galactose response. Our data indicate that rad53-11 GAL::SWI4-twas added 30 min prior to the release. The cultures were re-

cells do not enter into or progress through S phase anyleased from the arrest by shifting back to 25�. Aliquots of thesecultures were taken every 15 min and subjected to FACS analysis. faster than RAD GAL::SWI4-t cells. However, rad53-11 in

combination with Swi4-t results in a further increase inthe levels of both early (bubble) and late (stalled and

cells that overexpressed Swi4-t completed S phase sig- converged forks) rDNA replicative intermediates com-nificantly faster than controls, suggesting that Swi4-t may

pared to controls. rad53-11 also results in a significantincrease initiation in dbf4-1 cells (Figure 8B).

increase in the firing frequency of ARS608. This latterobservation is consistent with the previous report by

DISCUSSION Shirahige et al. (1998) made with a rad53-1 allele ofRAD53, but differs from the study of ARS301 in aSwi4-t is a hyperactive derivative of the late G1-specificrad53-21 strain where no increased firing was found (San-transcriptional activator Swi4. When overexpressed, ittocanale et al. 1999). This discrepancy is probably duecauses Swi4 target genes to be ectopically expressed atto differences between these RAD53 alleles. Only a subsetall stages of the cell cycle. In this study, we demonstratedof RAD53 mutations suppresses the temperature-sensi-that it can cause precocious entry into S phase. In addi-tive initiation defect of dbf4-1, and rad53-21 is syntheti-tion, we observed that even though GAL::SWI4-t cellscally lethal with dbf4-1 at the permissive temperaturecan enter S phase earlier than normal cells, they spend(Desany et al. 1998; Dohrmann et al. 1999; Kihara etmore time between the end of G1 and anaphase. Hemi-al. 2000).methylation analysis suggests that Swi4-t overexpression

We propose a simple interpretation of the data pre-leads to an extension of the time during which nascentsented in this study, which is that both GAL::SWI4-t andDNA is generated in a given region. If that is the case,rad53-11 affect initiation of replication, albeit to a differ-it could result from DNA reduplication or from gap fill-ent extent. In a normal cell, Swi4 activity is rate limitinging or strand break repair, or any combination thereof.for S-phase entry (McInerny et al. 1997). Elevating Swi42D gel analysis specifically suggests that Swi4-t overex-activity with a stable, constitutively produced form of Swi4pression correlates with a prolonged presence and/or in-may not only erroneously signal the cell to enter S, butcreased abundance of replicative intermediates. A highermay also signal its capacity to replicate at a high rate.level of replicative forks could be detected in rDNA inThis could lead to an increase in the number of replica-GAL::SWI4-t cells as compared to vector cells. Also, bub-tive forks concurrently present on DNA. When the den-ble and Y intermediates were detectable for a longer timesity of forks is elevated above what a given growth condi-in ARS608 DNA in GAL::SWI4-t cells.tion can support, this puts strain on the cell. ElevatedThe fact that we could detect an extended window of

time during which bubbles and Ys were present in rARS fork density could deplete dNTP and/or histone pools,


increase torsional stress on DNA, or make other key com- however, this locus is certainly not the only one that dif-fers between these strains. In mec1 and rad53 cells, Sml1-ponents of the replication machinery rate limiting. Themediated inhibition of the ribonucleotide reductase isconsequences of these adversities may be transient stall-not relieved (Zhao et al. 1998) and thus dNTP levelsing or slowing of fork movement.may be limiting. In the Vallen and Cross study (1999),Rad53 plays a critical role in mitigating the effects ofthe lethality of G1 cyclin overexpression to the mec1 mu-Swi4-t on replication. In a RAD cell, forks that stall ortants could be explained by a concomitant decline in theslow down may recruit Rad53. By stabilizing these forksmRNA level of the large subunit of the ribonucleotidewith the replicative machinery assembled on them, Rad53reductase Rnr1 (Vallen and Cross 1999), which fur-allows most of them to resume replication when suppliesther depletes dNTP pools. In our case, however, Swi4-tare replenished (Desany et al. 1998; Lopes et al. 2001).overproduction is lethal to rad53-11 cells even thoughStalled forks “marked” with activated Rad53 send a sig-it upregulates RNR1 transcription (J. Sidorova, unpub-nal inhibiting further origin firing and thus prevent fur-lished data), and SML1 is inactivated (Paulovich et al.ther accumulation of replicative intermediates and the1997). Hence, Rnr1 is not likely to be the cause of theunproductive spending of origins. Thus, the excessive fir-Swi4-t-induced lethality, even though it can mitigate theing promoted by GAL::SWI4-t is neutralized. In rad53-11lethal consequences of Swi4-t overproduction, when it,cells, this protective mechanism is deactivated, leadingtoo, is overexpressed. Moreover, suppression of the le-to the accumulation of many unstable replicative inter-thality of GAL::SWI4-t can also be achieved by overex-mediates, which may require disassembly and process-pression of SRL1 (J. Sidorova, unpublished data), anding through a recombination pathway to complete repli-thus it is not specific to RNR1. SRL1, whose function is un-cation (Rothstein et al. 2000). In agreement with this,known, is another Swi4-activated gene (Iyer et al. 2001),the defect in recombination caused by a rad52 disruptionand like RNR1, it can suppress the lethality of the rad53further exacerbates Swi4-t-induced lethality in rad53-11deletion when highly overexpressed (Desany et al. 1998).cells, but has no detectable effect on RAD53 Swi4-t cells.All these observations suggest that the Swi4-t-inducedThere is growing evidence that abundance of stalledlethality cannot be tracked to one Swi4-t target gene andforks during replication may contribute to genome desta-may instead result from altered expression of many targets.bilization. For example, mutations in the rDNA-specific

A dividing eukaryotic cell faces intrinsic challengeshelicase, Pif1, decrease the number of stalled forks, andthat arise from its ability to initiate DNA replicationthis is correlated with a decrease in the amount of rDNAfrom many loci throughout the genome. It seems thatbreakage and a reduction in the number of rDNA circlesfor an efficient S phase, it may be critically important to(Ivessa et al. 2000). Our work suggests the possibility thatadjust origin-firing efficiency depending on the cellularcells that experience precocious S phase due to a hyper-resources as well as to have a negative feedback con-activation of a G1-to-S transition regulator acquire geno-stantly monitoring the progress of replication.mic instability precisely because these regulators promote

We thank members of the Breeden lab for support and discussionsan increase in the number of stalled replicative forks.and Bonny Brewer for critical reading of the manuscript. Thanks areFinally, a number of hypotheses can be entertained todue to Steve Elledge, Andrew Emili, and Bonny Brewer for strainsexplain the mechanism by which hyperactive Swi4-t sig-and plasmids. This work was funded by grant GM-41073 from the

nals S-phase entry. Similar to other transcription factors, National Institutes of Health to L.B. J.S. was supported by the Leuke-a DNA-bound Swi4-t could promote initiation of replica- mia and Lymphoma Society Fellowship.tion in cis (Van de Vliet 1999). An interesting possibilityis raised by the recent report by Maser et al. (2001), whichfinds enhanced binding of mammalian E2F1 near active LITERATURE CITEDreplication origins. This and other findings (Royzman

Allen, J., Z. Zhou, S. Wolfram, E. C. Friedberg and S. J. Elledge,et al. 1999) imply that transcriptional activators impor-1994 The SAD1/RAD53 protein kinase controls multiple check-

tant for the G1-to-S transition could have a direct role points and DNA damage-induced transcription in yeast. GenesDev. 8: 2416–2428.in DNA replication control.

Andrews, B. J., and L. A. Moore, 1992 Interaction of the yeast Swi4It is also possible that the effect of Swi4-t on S-phase and Swi6 cell cycle regulatory proteins in vitro. Proc. Natl. Acad.progression is mediated by the altered expression of Swi4 Sci. USA 89: 11852–11856.

Baetz, K., and B. Andrews, 1999 Regulation of cell cycle transcrip-target genes. Two critical targets of the Swi4 activator,tion factor Swi4 through auto-inhibition of DNA binding. Mol.which are rate limiting for the G1-to-S transition, are the Cell. Biol. 19: 6729–6741.

G1 cylins CLN1 and CLN2. Overproduction of CLN1 or Bashkirov, V. I., J. S. King, E. V. Bashkirova, J. Schmuckli-Maurerand W.-D. Heyer, 2000 DNA repair protein Rad55 is a terminalCLN2 is lethal in combination with mec1 and, to a lessersubstrate of the DNA damage checkpoints. Mol. Cell. Biol. 20:degree, rad53 mutations in some backgrounds (Vallen 4393–4404.

and Cross 1995, 1999). However, CLN overproduction Bousset, K., and J. Diffley, 1998 The Cdc7 protein kinase is re-quired for origin firing during S phase. Genes Dev. 12: 480–490.is not sufficient to kill either mec1-1 or rad53-11 cells in

Breeden, L., and G. Mikesell, 1991 Cell cycle-specific expressionthe A364a background. This difference may be attrib- of the SWI4 transcription factor is required for the cell cycle

regulation of HO transcription. Genes Dev. 5: 1183–1190.uted to the state of the SML1 locus (Zhao et al. 1998);


Breeden, L., and G. Mikesell, 1994 Three independent forms of element in the SWI4, CLN3, CDC6 and CDC47 promoters activatesregulation affect expression of HO, CLN1 and CLN2 during the M/G1-specific transcription. Genes Dev. 11: 1277–1288.cell cycle of Saccharomyces cerevisiae. Genetics 138: 1015–1024. Mendenhall, M., and A. Hodge, 1998 Regulation of Cdc28 cyclin-

Brewer, B. J., D. Lockshon and W. L. Fangman, 1992 The arrest dependent protein kinase activity during the cell cycle of the yeastof replication forks in the rDNA of yeast occurs independently Saccharomyces cerevisiae. Microbiol. Mol. Biol. Rev. 62: 1191–1243.of transcription. Cell 71: 267–276. Nasmyth, K., and L. Dirick, 1991 The role of SWI4 and SWI6 in

Desany, B. A., A. A. Alcasabas, J. B. Bachant and S. J. Elledge, the activity of G1 cyclins in yeast. Cell 66: 995–1013.1998 Recovery from DNA replicational stress is the essential Navas, T. A., Y. Sanchez and S. J. Elledge, 1996 RAD9 and DNAfunction of the S-phase checkpoint pathway. Genes Dev. 12: 2956– polymerase £ form parallel sensory branches for transducing the2970. DNA damage checkpoint signal in Saccharomyces cerevisiae. Genes

Dohrmann, P. R., G. Oshiro, M. Tecklenburg and R. A. Sclafani, Dev. 10: 2632–2643.1999 RAD53 regulates DBF4 independently of checkpoint func- Nugroho, T. T., and M. D. Mendenhall, 1994 An inhibitor oftion in Saccharomyces cerevisiae. Genetics 151: 965–977. yeast cyclin-dependent protein kinase plays an important role in

Donaldson, A. D., W. L. Fangman and B. J. Brewer, 1998 Cdc7 ensuring the genomic integrity of daughter cells. Mol. Cell. Biol.is required throughout the yeast S phase to activate replication 14: 3320–3328.origins. Genes Dev. 12: 491–501. Ogas, J., B. J. Andrews and I. Herskowitz, 1991 Transcriptional

Donehower, L. S., 1997 Genetic instability in animal tumorigenesis activation of CLN1, CLN2, and a putative new G1 cyclin (HCS26)models. Cancer Surv. 29: 329–352. by SWI4, a positive regulator of G1-specific transcription. Cell

Durocher, D., J. Henckel, A. R. Fersht and S. P. Jackson, 1999 66: 1015–1026.The FHA domain is a modular phosphopeptide recognition mo- Paulovich, A. G., and L. H. Hartwell, 1995 A checkpoint regu-tif. Mol. Cell 4: 387–394. lates the rate of progression through S Phase in S. cerevisiae in

Emili, A., 1998 MEC1-dependent phosphorylation of Rad9p in re- response to DNA damage. Cell 82: 841–847.sponse to DNA damage. Mol. Cell 2: 183–189. Paulovich, A. G., R. U. Margulies, B. M. Garvik and L. H. Hart-

Epstein, C. B., and F. R. Cross, 1994 Genes that can bypass the well, 1997 RAD9, RAD17, and RAD24 are required for S phaseCLN requirement for Saccharomyces cerevisiae cell cycle START. regulation in Saccharomyces cerevisiae in response to DNA damage.Mol. Cell. Biol. 14: 2041–2047. Genetics 145: 45–62.

Felsher, D. W., and J. M. Bishop, 1999 Transient excess of MYC Pellicioli, A., C. Lucca, G. Liberi, F. Marini, M. Lopes et al., 1999activity can elicit genomic instability and tumorigenesis. Proc. Activation of Rad53 kinase in response to DNA damage and itsNatl. Acad. Sci. USA 96: 3940–3944. effect in modulating phosphorylation of the lagging strand DNA

Foiani, M., A. Pellicioli, M. Lopes, C. Lucca, M. Ferrari et al., polymerase. EMBO J. 18: 6561–6572.2000 DNA damage checkpoints and DNA replication controls Primig, M., S. Sockanathan, H. Auer and K. Nasmyth, 1992 Anat-in Saccharomyces cerevisiae. Mutat. Res. 451: 187–196. omy of a transcription factor important for the start of the cell

Friedman, K. L., and B. J. Brewer, 1995 Analyis of Replication Inter- cycle in Saccharomyces cerevisiae. Nature 358: 593–597.mediates by Two-Dimensional Agarose Gel Electrophoresis. Academic Rose, M. D., F. Winston and P. Hieter, 1990 Methods in Yeast Ge-Press, San Diego. netics. Cold Spring Harbor Laboratory Press, Cold Spring Har-

Friedman, K. L., M. K. Raghuraman, W. L. Fangman and B. J. bor, NY.Brewer, 1995 Analysis of the temporal program of replication Rothstein, R., M. Benedicte, and S. Gangloff, 2000 Replicationinitation in yeast chromosomes. J. Cell Sci. 19(Suppl.): 51–58. fork pausing and recombination or “gimme a break.” Genes Dev.

Friedman, K. L., B. J. Brewer and W. L. Fangman, 1997 Replication 14: 1–10.profile of Saccharomyces cerevisiae chromosome VI. Genes Cells Royzman, I., R. J. Austin, G. Bosco, S. P. Bell and T. L. Orr-2: 667–678. Weaver, 1999 ORC localization in Drosophila follicle cells and

Huang, M., Z. Zhou and S. J. Elledge, 1998 The DNA replication the effects of mutations in dE2F and dDP. Genes Dev. 13: 827–840.and damage checkpoint pathways induce transcription by inhibi- Sanchez, Y., J. Bachant, H. Wang, F. Hu, D. Liu et al., 1999 Controltion of the Crt1 repressor. Cell 94: 595–605. of the DNA damage checkpoint by Chk1 and Rad53 protein

Ivessa, A. S., J.-Q. Zhou and V. A. Zakian, 2000 The Saccharomyces kinases through distinct mechanisms. Science 286: 1166–1171.Pif1p DNA helicase and the highly related Rrm3p have opposite Santocanale, C., and J. F. X. Diffley, 1998 A Mec1-0 and Rad53-effects on replication fork progression in ribosomal DNA. Cell dependent checkpoint controls late-firing origins of DNA replica-100: 479–489. tion. Nature 395: 615–618.Iyer, V. R., C. E. Horak, C. S. Scafe, D. Botstein, M. Snyder et al.,Santocanale, C., K. Sharma and J. F. X. Diffley, 1999 Activation2001 Genomic binding sites of the yeast cell-cycle transcription

of dormant origins of DNA replication in budding yeast. Genesfactors SBF and MBF. Nature 409: 533–538.Dev. 13: 2360–2364.Jacobs, C. W., A. E. M. Adams, P. J. Szaniszlo and J. R. Pringle,

Schwob, E., T. Bohm, M. D. Mendenhall and K. Nasmyth, 19941988 Functions of microtubules in the Saccharomyces cerevis-The B-type cyclin kinase inhibitor p40SIC1 controls the G1 to Siae cell cycle. J. Cell Biol. 107: 1409–1426.transition in S. cerevisiae. Cell 79: 233–244.Johnston, L. H., and A. P. Thomas, 1982 A further two mutants

Shirahige, K., Y. Hori, K. Shiraishi, M. Yamashita, K. Takahashidefective in initiation of the S phase in the yeast Saccharomyceset al., 1998 Regulation of DNA-replication origins during cell-cerevisiae. Mol. Gen. Genet. 186: 445–448.cycle progression. Nature 395: 618–621.Kihara, M., W. Nakai, S. Asano, A. Suzuki, K. Kitada et al., 2000

Sidorova, J., and L. Breeden, 1993 Analysis of the SWI4/SWI6 pro-Characterization of the yeast Cdc7p/Dbf4p complex purifiedtein complex, which directs G1/S-specific transcription in Saccha-from insect cells. Its protein kinase activity is regulated by rad53p.romyces cerevisiae. Mol. Cell. Biol. 13: 1069–1077.J. Biol. Chem. 275: 35051–35062.

Sidorova, J., and L. L. Breeden, 1997 Rad53-dependent phosphory-Linskens, M. H. K., and J. A. Huberman, 1988 Organization oflation of Swi6 and down-regulation of CLN1 and CLN2 transcrip-replication of ribosomal DNA in Saccharomyces cerevisiae. Mol.tion occur in response to DNA damage in Saccharomyces cerevisiae.Cell. Biol. 8: 4927–4935.Genes Dev. 11: 3032–3045.Lopes, M., C. Cotta-Ramusino, A. Pellicioli, G. Liberi, P. Plevani

Spruck, C. H., K. A. Won and S. I. Reed, 1999 Deregulated cyclinet al., 2001 The DNA replication checkpoint response stabilizesE induces chromosome instability. Nature 401: 297–300.stalled replication forks. Nature 412: 557–561.

Sun, Z., D. S. Fay, F. Marini, M. Foiani, and D. F. Stern, 1996 Spk1/Mai, S., J. Hanley-Hyde, G. J. Rainey, T. I. Kuschak, J. T. Paul etRad53 is regulated by Mec1-dependent protein phosphorylational., 1999 Chromosomal and extrachromosomal instability of thein DNA replication and damage checkpoint pathways. Genes Dev.cyclin D2 gene is induced by Myc overexpression. Neoplasia 1:10: 395–406.241–252.

Sun, Z., J. Hsiao, D. S. Fay and D. F. Stern, 1998 Rad53 FHAMaser, R. S., O. K. Mirzoeva, J. Wells, H. Olivares, B. R. Williamsdomain associated with phosphorylated Rad9 in the DNA damageet al., 2001 Mre11 complex and DNA replication: linkage tocheckpoint. Science 281: 272–274.E2F and sites of DNA synthesis. Mol. Cell. Biol. 21: 6006–6016.

Sung, P., K. M. Trujillo and S. VanKomen, 2000 RecombinationMcInerny, C. J., J. F. Partridge, G. E. Mikesell, D. P. Creemerand L. L. Breeden, 1997 A novel Mcm1-dependent promoter factors of Saccharomyces cerevisiae. Mutat. Res. 451: 257–275.


Tersero, J. A., and J. F. X. Diffley, 2001 Regulation of DNA replica- Tel1-dependent hyperphosphorylation and interacts with Rad53after DNA damage. EMBO J. 17: 5679–5688.tion fork progression through damaged DNA by the Mec1/Rad53

checkpoint. Nature 412: 553–557. Weinert, T., 1998 DNA damage and checkpoint pathways: molecu-lar anatomy and interactions with repair. Cell 94: 555–558.Tyers, M., G. Tokiwa and B. Futcher, 1993 Comparison of the

Saccharomyces cerevisiae G1 cyclins: Cln3 may be an upstream activa- Yamashita, M., Y. Hori, T. Shinomiya, C. Obuse, T. Tsurimoto etal., 1997 The efficiency and timing of initation of replicationtor of Cln1, Cln2 and other cyclins. EMBO J. 12: 1955–1968.

Vallen, E. A., and F. R. Cross, 1995 Mutations in RAD27 define a of multiple replicons of Saccharomyces cerevisiae chromsomepotential link between G1 cyclins and DNA replication. Mol. Cell. VI. Genes Cells 2: 655–665.Biol. 15: 4291–4302. Zhao, X. L., E. G. D. Muller and R. A. Rothstein, 1998 A suppres-

Vallen, E. A., and F. R. Cross, 1999 Interaction between the MEC1- sor of two essential checkpoint genes identifies a novel proteindependent DNA synthesis checkpoint and G1 cyclin function in that negatively affects dNTP pools. Mol. Cell 2: 329–340.Saccaromyces cerevisiae. Genetics 151: 459–471. Zheng, P., D. S. Fay, J. Burton, H. Xiao, J. L. Pinkham et al., 1993

Van de Vliet, P. C., 1999 Roles of Transcription Factors in DNA Repli- SPK1 is an essential S-phase specific gene of Saccharomyces cerevisiaecation. Cold Spring Harbor Laboratory Press, Cold Spring Har- that encodes a nuclear serine/threonine/tyrosine kinase. Mol.bor, NY. Cell. Biol. 13: 5829–5842.

Vialard, J. E., C. S. Gilbert, C. M. Green and N. F. Lowndes, 1998The budding yeast Rad9 checkpoint protein is subjected to Mec1/ Communicating editor: P. Russell

precocious s-phase entry in budding yeast prolongs replicative … · copyright 2002 by the...

Documents